Duke Wiki  logo
Child pages
  • ENV 255 - Applied Regression Analy.
Skip to end of metadata
Go to start of metadata

Created/Maintained by NSSC,
with the support of
Nicholas Marketing
and Communications
and Nicholas IT.

Need help? Email Nicki Admin

Content within this site
do NOT represent the views
or endorsements of the
Nicholas School and
Duke University. Opinions
herein are solely those
of contributors.

Insert name of course

Course Stats

Instructor(s):  Song Qian Ph.D

Units:3

Semester Offered: Spring

Description

(Official Nicholas School course description link)

It is as the name implies, a course in "applied" statistical analysis.  It is geared towards students who will be using statistics in their Master's Projects and need guidance.  There is not a set syllabus or curriculum for the course, instead the professor designs the course around the specific types of statistical analysis that the class participants will be using for their data.  Lecture time is spent between the professor's presentation of the class selected topics, and regular presentations by the students discussing their ongoing progress with the analysis of their data.  Students get experience using applied statistics on thier own data set, while also getting exposed to various other statistical applications that their classmates are investigating.

Skills and Career Applications

(Insert comments on the intent and career applications of this class. Comments/Quotes from Nicholas Alumni are especially valued.)

 This class will be useful to anyone who will be using statistics in their future careers.

Registration Advice

(Insert comments here on how prospective students should approach registration for this class, e.g. hard to get into; sign up early; permission numbers; don't take if you already have heavy load)

Student Opinions

(This space is intended to compile general comments from past students on the course. It will be judiciously moderated for extreme and unprofessional comments, but otherwise, it is open to a broad range of student feedback.)

 This class was extremely helpful for my Master's Project and my understanding of statistics in general.  I highly recommend this class to anyone who intends to use statistics in their master's project.  This is for two reasons, 1) the only assignments in the class are to work on your data set and occasionally present to the class your current progress, findings, or problems.  This means that if you are working on your MP dataset, you are getting guidance and advice while working towards the completion of your MP Analysis.  2)  The professor is very helpful, and has a vast knowledge of many different types of statistics.  For my project i was interested in a geostatistical analysis, which at the start i neither understood nor knew of anyone who understood.  The intructor was very helpful in guiding me through the analysis and pointing me in the right direction.  Bonus- this class makes statistics much more interesting, because you see how it can actually be applied to real data sets.  Bottom line - Don't let the title of this course or the fact that you may not have loved Env.210 deter you from statistics, take this class and you will not regret it.

The Instructor's Take...

(This space is exclusively for the instructor to broadly comment on his or her course, and respond to commonly received feedback, explain methods and approaches, and encourage student registration. Instructors: Please limit to 300 words, or link to a wiki page of your own creation to explain in detail.)

This course was designed to teach regression.  However, with the use of computer software, many topics originally included in the course is no longer necessary.  As I learned from students in the past few years, the most helpful quantitative class is one that targeted specifically to students' course/MP work.  Instead of teaching regression and generalized linear models, students are now asked to use their master's projects as the basis to form their class projects and learn the necessary materials to complete their projects.  During spring of 2009, five MEM students and two PhD students participated the class.  Based on students interests, the class covered four topics (regression with correlated residuals -- including spatial data and time-series data, multilevel regression, probability density estimation for delineating animal habitats, and multinomial regression for species composition data).  If your MP has a quantitative component, you should consider this course.  I am always looking for interesting applications of many interesting statistical techniques.

  • (Insert links/attachments for past syllabi, textbooks, etc.)
  •